Numpy implementation of Popular Algorithms and Architecture of Deep Neural Networks.

Initilization Methods -

1. Random initilization
2. He initilization

Forward Propogation -

1. Forward Propogation without Batch Normalization
2. Forward Propogation with Batch Normalization

Activation Functions -

1. Sigmoid
2. ReLU
3. Leaky-ReLU

Regularization -

1. L-2 regularization
2. Dropout regularization

Backpropogation -

1. Backpropogation without Batch Normalization
2. Backpropogation with Batch Normalization

Cost -

1. Negative Log Likelihood Cost function.

Debugging -

1. Gradient Checking